[prev in list] [next in list] [prev in thread] [next in thread] 

List:       mysql
Subject:    Re: MySQL Benchmark. OT and beyond...
From:       RV Tec <rvt () shell2 ! lomag ! net>
Date:       2004-05-31 13:18:19
Message-ID: Pine.BSF.4.58.0405310907460.29064 () shell2 ! lomag ! net
[Download RAW message or body]

Folks,

> FreeBSD 5.x does a lot of things really well, and it can be very
> fast too, but there's no stable release of 5.x yet, and the 5.x
> code is clearly hampered speedwise by the presence of loads of
> debugging code. From the 5.x documentation:
>
> ---
> NOTE TO PEOPLE WHO THINK THAT FreeBSD 5.x IS SLOW:
>          FreeBSD 5.x has many debugging features turned on, in
>          both the kernel and userland.  These features attempt to detect
>          incorrect use of system primitives, and encourage loud failure
>          through extra sanity checking and fail stop semantics.  They
>          also substantially impact system performance.  If you want to
>          do performance measurement, benchmarking, and optimization,
>          you'll want to turn them off.  This includes various WITNESS-
>          related kernel options, INVARIANTS, malloc debugging flags
>          in userland, and various verbose features in the kernel.  Many
>          developers choose to disable these features on build machines
>          to maximize performance.
> ---
>
> If Linux is really twice as fast as FreeBSD, as people have reported
> recently, then go ahead and use it if that's what you want. But I
> hope everyone reporting this is using optimized software, and isn't
> complaining about a FreeBSD that's explicitly slowed down by running
> under gobs of debugging code.
>
> So I would imagine you'd see a huge speedbump from "downgrading" to 4.10.

If it makes any sense, I did turn off everything related to DEBUG at the
kernel. I also removed support of *TRACE. Also did a make world before
compiling MySQL (thus, before the trials).

It was really strange to see that kind of performance, when the expected
was a lot higher. I can't remember if I shared the results with MySQL
4.0.20 running on a FreeBSD 5.2.1 with pthreads, but they weren't that
different from LinuxThreads.

As I write this e-mail, I'm configuring a Linux (Debian) box, later on
I'll share those results.

Best regards,
RV Tec

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:    http://lists.mysql.com/mysql?unsub=mysql@progressive-comp.com

[prev in list] [next in list] [prev in thread] [next in thread]